Search

The objective of this study was to investigate the impact of flow, transport and geochemical parameters in the unsaturated and saturated zones on the release of SO4 from overburden lignite spoil piles into the adjacent lake. A vertical one-dimensional model was set up using the reactive transport simulator SULFIDOX in order to account for the unsaturated zone. The SULFIDOX model was calibrated for effective diffusion using measured O2 in the gas phase and SO4 concentrations in the liquid phase from the unsaturated zone of the heap. The results show high sensitivity to O2 supply and initially present gypsum, but the inclusion of secondary mineral precipitation in equilibrium is of minor importance for the results. To account for the transport of released SO4 from the saturated zone into the surface water, scenarios were performed by using SULFIDOX results as input concentration for a two-dimensional vertical model set up with PROCESSING MODFLOW and MT3D. These scenarios indicate a rising discharge of SO4 into the adjacent lake due to continued pyrite weathering for 80 a. Results are highly sensitive to dispersivity, whereas the spatial variability of pyrite distribution did not show any influence on the results. The consideration of initially present gypsum shows a major effect on the modelled SO4 release.

The recently developed geochemical modelling code, SULFIDOX, has been applied to simulate weathering of a waste rock dump at the Aitik mine site, Sweden. SULFIDOX models the key chemical and physical processes in the dump temporally and spatially (in two dimensions). The following processes are represented: gas and heat transport; water infiltration; aqueous speciation; mineral dissolution/oxidation and precipitation. Field observations at the site suggest that sulphide oxidation rates within the dump are variable. Although the major part of the dump is oxidising slowly, there are pockets of more highly oxidising material, particularly toward the dump edges. Using SULFIDOX, several models of the dump were investigated: (i) a dump wholly comprised of slowly oxidising material (representing a case where water flow paths are such that no rapidly oxidising regions are accessed); (ii) a dump wholly comprised of the more rapidly oxidising material (representing the opposite (and probably unlikely) extreme, where water flows only through rapidly oxidising regions in the dump); and (iii) a dump comprising a mixture of both slowly and more rapidly oxidising material, that more closely represents the mix of material in the dump. All the models studied gave O2 depth profiles consistent with those observed in probe holes at the site, and confirmed that only a minimal amount of heat production would be expected in the dump due to the role of exothermic sulphide oxidation reactions. The models suggested that a medium-term steady-state, with respect to effluent chemistry, would be achieved after 3–4 years. Based on sulphide consumption rates during this steady-state period, the time periods required to consume all the sulphide in the dump range from a few hundred to many thousands of years. Using the mixed model, and based on a mixture containing 86% slowly and 14% rapidly oxidising material, the calculated effluent chemistry was in good agreement with the observed effluent chemistry. Improvements with respect to the K concentrations were possible by including precipitation of a K-bearing secondary mineral such a K-jarosite in the model. Results from the more rapidly oxidising model suggested that gypsum precipitation might be expected in those regions of the dump containing this material. In summary, the SULFIDOX modelling code has been used successfully to reproduce observed data for the Aitik waste-rock dump. Using SULFIDOX, valuable insight was gained in relation to the temporal and spatial evolution of the dump.